Improving Water Productivity-new Avenues (IWMI-TATA)

7/31/2019 Improving Water Productivity-new Avenues (IWMI-TATA)

1/17

185

IMPROVING WATER PRODUCTIVITY IN AGRICULTURE IN DEVELOPING

ECONOMIES: IN SEARCH OF NEW AVENUES

M. Dinesh Kumar and Jos C. van Dam1

Abstract

This article shows how the various considerations for analyzing water productivity (WP) differ due to the

differences in stakeholder interests, and objectives and units of analysis. Also it identifies some major gaps in WP

research and the key drivers of change in WP. The main arguments are: 1] in developing economies like India the

objective of WP research should also be to maximize net return per unit of water and aggregate returns for the

farmer, rather than merely enhancing crop per drop; 2] the determinant for analyzing the impact of efficien

irrigation technologies on basin level WP and water saving should be consumed fraction (CF) rather than evapo-

transpiration; 3] in closed basins, determinants for analyzing basin level WP improvement through water harvesting

and conservation should be incremental economic returns & opportunity costs; 4] at the field level, the reliabilityof irrigation water and changing water allocation could be the key drivers of change in WP that need to be

analyzed, whereas at the farm level, changes in the crop mix and farming system could be key drivers of change. In

composite farming systems, measures to enhance WP should be based on farm-level analysis involving considerations

such as risk taking ability and investment capabilities of the farmer. Finally, the options to enhance WP in agriculture

seem to be quite limited, given the larger objective of addressing food security, poverty alleviation, and employmen

generation concerns in rural areas.

1. INTRODUCTION

Water productivity in agriculture would be the single most important factor driving the water use

globally in the future (Molden et al., 2000; Rijsberman, 2004). Hence, research to evaluate crop water productivityand analyze the drivers of change in the same, has fascinated many researchers and scholars worldwide(Ahmad et al., 2004; Ambast et al., 2006; Grismer, 2001; Howell, 2001; Kijne et al., 2003; Zwart and Bastiaanssen2004; Singh, 2005; van Dam et al., 2006). As a result, most of the research studies on crop water productivitywere undertaken in naturally water-scarce regions of the world. Such regions include western United Statesdrought-prone areas of arid Australia, semi arid areas of Indian and Pakistan Punjab, Turkey, and Mexico.

Water productivity in crop production can be expressed in terms of biomass production per cubicmetre of water diverted or depleted (kg/m3), known as physical productivity of water; and net or gross presenvalue of the crop produced per cubic metre of water diverted or depleted (Rs/m3) known as economic productivityof water (Kijne et al., 2003). A recent synthesis by Zwart and Bastiaanssen (2004) of an extensive body ofliterature available research world over showed that water productivity in terms of biomass output per unit odepleted water (kg/ET) or physical productivity of water in crop production has been mostly analyzed acrossthe world at least for some of the major crops; and enough is already known about the factors that explain itsvariations across locations. But, it also showed that no attention is paid to know how the crops compare interms of economic returns from every unit of water depleted. But, this is crucial, because the measures toenhance water productivity of a crop such as higher dosage of nitrogenous fertilizers; improved soil managementbetter agronomic practices, including the use of high yielding varieties, and pest control; water harvesting andsupplementary irrigation; and investment in water delivery control measures, have economic imperatives.

1 Researcher and ITP Leader, IWMI, South Asia Sub-regional Office, Hyderabad, and Associate Professor and Chair-Soil, Agrohydrology and Groundwater Management, Faculty of Environmental Sciences, Wageningen University and Research CentreWageningen, the Netherlands, respectively. Email: [email protected]


2/17

186

This heavy focus on physical productivity of water is perhaps because most of these analyses weredone by agricultural scientists, who are concerned with raising the dry matter yield of crop per unit of evapo-transpiration. The other factors, which might have been responsible for this bias are: 1] water is a limitingfactor at the societal level for enhancing crop production in these regions (Howell, 2001: pp), which still havelarge cropped areas under un-irrigated conditions (Loomis and Connor, 1996: pp10), and water productivityimprovement enable farmers to divert part of the saved water to expand irrigated area; and, 2] with volumetricrationing and prices farmers have to pay for water, they are likely to get higher net returns along with higheyields through efficient irrigation technologies that reduce consumptive use2. Another factor could be thefluctuating price of agricultural commodities in the market, which changes the net return per unit volume ofwater.

But, the avenues to improve agricultural WP through farming system changes are not explored. This ia major shortcoming, when we consider the fact that most of the farms in developing economies like India andmost of Africa are complex with several crops; and also composite with crops and dairying instead of one otwo crops. After Rothenberg (1980), as farms are organized to maximize the net economic return they are thebest fundamental units for economic analysis. Hence, how productively farmers use their water cannot beassessed in relation to particular crop alone, but in relation to the entire farm. In sum, this dominant paradigmof more crop per drop influence WP research in Asia and Africa.

On the other hand, there has also been greater recognition of the distinction between securing fieldlevel water-saving and field-level WP improvement, and water-saving and WP improvement at the basin-scale(Allan et al., 1998; Howell, 2001; Molle and Turral, 2004; Seckler et al., 2003). The concept of open basinsand closed basins is often used to explain how the determinants of WP could be manipulated and water savingachieved, or otherwise, in different situations. The received wisdom is that in closed basins, field-level watesaving does not result in water-saving and WP improvement at the basin level, except when the return flowsmeet with saline aquifers or are non-returnable; and otherwise basin level water saving and WP improvementcomes only from reduction in consumptive use (Molle and Turral, 2004).

This new paradigms in water resource management also seems to have influenced research in manycountries in Asia and Africa: 1] in deciding what one should look for as key determinants in WP analysis; and2] in identifying the drivers of change in WP. They have hardly captured the complex technical, social, economicinstitutional and policy settings that govern water allocation policies by government and water use decisions byfarmers. This concerns the poor technical efficiency and reliability of public canal systems; heavily subsidies inpricing of water and electricity in farm sector; huge public investments in water harvesting; and, lack oinstitutional regimes governing the use of water from canal schemes and groundwater.

This article first takes a critical look at these two paradigms in agricultural water management tosee how far they are useful in exploring new avenues for WP improvements and water saving, particularlyin situations like India. It also explores new opportunities for WP improvements and water saving for fieldsfarms and regions, by analyzing the complex variables which drive these WP parameters, and identifies newareas for research.

The questions being addressed are as follows. 1. Given the heavy subsidies in electricity and water

used for agriculture and lack of well-defined rights in surface water and groundwater in developing countrielike India, does research on raising crop per drop make sense, or what should be the new determinants of WPfor both farmer and basin water managers? 2. What considerations should be involved in analyzing basin leveWP and water saving impacts of efficient irrigation? 3. What are the likely impacts of improved reliability oirrigation, and changing water allocation on crop water productivity and water saving? 4. In composite farmingsystems, what should be key objectives and priority areas of WP research? 5. What should be the priority areafor research on enhancing regional WP in agriculture, in countries like India where food security, rural employmenand poverty alleviation are still major issues?

2 As water saving leads to cost saving in irrigation sufficient to offset the additional cost of fertilizer and technology inputs.


3/17

187

2. WHY A NEW PARADIGM OF RESEARCH ON AGRICULTURAL WATER PRODUCTIVITY

IN INDIA?

2.1 More Income Returns Vs. More Crop Per Drop

The main considerations involved in analyzing WP in the West is in reducing the amount of waterequired to produce a unit weight of crop, as this would automatically ensure higher net return per unit of landBut this is not the concern in many developing economies in Asia, where land use intensity is already very high

in many regions. Surface water is heavily subsidized, and pricing is also inefficient (Kumar, 2003). There iszero marginal cost of electricity used for pumping groundwater for irrigation (Kumar, 2005). Hence, the measuresto enhance water productivity through ET reduction and yield enhancement may not result in significanimprovement in net income for the farmer for a unit area of irrigated land, though net water productivity inrupee terms may increase. While major investments are required to achieve irrigation efficiency improvementand yield enhancement, the increased benefit farmers get is only in terms of market price for higher yield. Thereason is that the real water-saving and energy saving3, which are major impacts of the technological interventionsdo not get converted into saving in private costs of water.

A study by Sander Zwart (2006), which involved analysis of system level WP in irrigated wheat in sixdifferent regions around the world using SEBAL (Surface Energy Balance) methodology, shows that the variationin WP is not so much due to variations in ET, but due to variations in yield (see Table 1). The average ET wahighest in Pakistan (443mm) and lowest in Sirsa (361mm), which is approximately 10% higher/lower than theaverage (source: analysis by Sander J. Zwart, 2006). Though the potential evapo-transpiration (PET) dependson the climate, especially the relative humidity (air temperature and solar radiations remaining in a narrow rangeacross these six regions), actual ET could have been manipulated by changing water available to crops throughirrigation. But, this does not seem to have happened. As a consequence, WP is strongly related to wheat yieldsThe reason that ET remains the same is that there is a shift from evaporation (E) to transpiration (T). As soonas the environment for crop production are improved (fertilizers, weeding, better seeds, water managementetc., etc.) there will be a shift from non-beneficial to beneficial water depletion. This shows enhancement in WP(kg/ET) can mainly come from crop technologies, which needs farmer investments.

3 Whether use of efficient irrigation technologies can reduce energy use for irrigation or increase depends on the type of irrigationtechnology and how pressurized is the traditional water supply (Loomis and Connors, 1996).

Table 1: Average System-level Water Productivity in Wheat in six Different Wheat GrowingRegions around the World

Nile Delta, Egypt 408 (59) 6.1 (0.9) 1.50 (0.12)

Yaqui Valley, Mexico 402 (36) 5.5 (0.9) 1.37 (0.16)

Sirsa, India 361 (16) 4.4 (0.3) 1.22 (0.06)

Linxian County, China 436 (35) 3.8 (1.4) 0.86 (0.28)Hebei Province, China 380 (50) 2.5 (0.9) 0.64 (0.21)

Sindh Province, Pakistan 443 (82) 2.2 (0.7) 0.50 (0.11)

Source: analysis by Sander J. Zwart dated May, 2006

LocationAverage ET/

StandardDeviation (mm)

Average yield(ton ha-1)

Average WPET

(kg m-3)

Now, the only way to create incentive among farmers to adopt efficient irrigation technologies for WPimprovement is to subsidize it. The idea is to make private benefits offset the private costs (Kumar, 2007)


4/17

188

While yield enhancement is also a benefit of efficient irrigation technologies (Loomis and Connor, 1996: pp398)it can also come from improved agronomic practices mentioned above. The extent of subsidy for a systemwhich can save X amount of water could be kept higher than the difference between the private costs andbenefits. It should be guided by the positive externality that X creates on the society. Since, governmensubsidies for efficient irrigation technologies are extremely limited in developing countries 4 such measures toenhance WP do not result in increased land productivity.

This means that they have to divert part of the water saved to another plot to sustain their income asnet return is WP multiplied by the volume of water. But, in situations where the entire holding is used, farmerwill not have much incentive to go for measures that do not increase their returns from the land, but onlyreturns per unit of water. This is the situation in India, where the average holding of farmers is quite low (lessthan 1 ha) when compared to that in Western US or Australia. The size of median landholding in Australia is 300ha (ABS, 2002). This clearly means that what is socially optimal is that farmers look for alternatives thaenhance productivity of their land remarkably, simultaneously reducing water requirement, or divert part of thewater to other water-based farming systems that have minimal dependence on land. In nutshell, there is a cleatrade off between enhancing physical productivity of water, and maximizing income returns. This argumentalso holds true when it comes to analyzing the WP impacts of water harvesting for supplementary irrigationwhich happens with public investment. This is dealt with the subsequent section.

2.2 Poor Focus on Economics of Water Harvesting and Supplementary Irrigation

In the west, the focus in WP research has been on efficient irrigation technologies, including those fosupplementary irrigation, in some African countries (Oweis et al., 1999; Rockstrm et al., 2002), Mexico(Scott and Silva-Ochoa, 2001) and in India, the focus has shifted to potential impact of water harvesting.

This is applicable to some of the recent work in eastern African countries. Rockstrm et al., (2002)have shown remarkable effect of supplementary irrigation through water harvesting on physical productivity owater expressed in kg/ET, for crops as sorghum and maize. However, the research did not evaluate the incrementaeconomic returns due to supplementary irrigation against the incremental costs of water harvesting. It also doesnot quantify the real hydrological opportunities available for water harvesting at the farm level and its reliabilityThe work by Scott and Silva-Ochoa (2001) in the Lerma-Chapala basin in Mexico showed higher gross value

product from crop production in areas with better allocation of water from water harvesting irrigation systemsBut, their figures of surplus value product which takes into account the cost of irrigation are not available fromtheir analysis. In arid and semi arid regions, the hydrological and economic opportunities of water harvestingare often over-played. A recent work in India has shown that the cost of water harvesting systems would beenormous, and reliability of supplies from it very poor in arid and semi arid regions of India, which are characterizedby low mean annual rainfalls, very few rainy days, high inter-annual variability in rainfall and rainy days, andhigh potential evaporation leading to a much higher variability in runoff between good rainfall years and poorrainfall years (Kumar et al., 2006).

With high capital cost of WH systems needed for supplemental irrigation, the small and marginal farmers would have less incentive to go for it. The reason is incremental returns due to yield benefits may

not exceed the cost of the system. This is particularly so for crops having low economic value such as wheaand paddy, which dominate arid and semi arid regions in India. But, even if the benefits due to supplementaryirrigation from water harvesting exceed the costs, it will not result in higher WP in economic terms in closedbasins. The exception is when the incremental returns are disproportionately higher than the increase in ETThis is because, in a closed basin, increase in beneficial ET at the place of water harvesting will eventuallyreduce the beneficial use d/s. Lack of this economic perspective in decisions, however, results in too much

4 For instance, the government of India had provided Rs. 5 billion towards subsidy for drip and sprinkler systems in the five yearplan. But, this amount is just sufficient to cover an area of 100,000 ha against a total net irrigated area of nearly 55 m haaccounting for just 0.20%, if one considers an investment of Rs. 100000 per ha of area under MI system, and a subsidy tothe tune of 50%.


5/17


6/17

190

yield, and in spite of higher cost of irrigation higher net returns as compared to canal irrigators (Kumar andSingh, 2001; IRMA/UNICEF, 2001) there is limited research data on the differential productivity of groundwateirrigation over surface irrigation. A recently published study for the Andalusian region (Southern Spain) showsthat each cubic meter of groundwater used for irrigation provides five times more money and almost four timesmore jobs than a cubic meter of surface water used also for irrigation (Hernndez-Mora et al., (1999).

But, how this positive differential reliability in case of well irrigation does get translated into WP gainis a major point of enquiry. There are two possibilities. First, it is an established fact that while crop yieldincreases in proportion to increase in transpiration, at higher doses irrigation does not result in beneficia

transpiration, but non-beneficial evaporation. Irrigation water dosages are normally higher in canal irrigationThis way, increased CF does not result in proportional increase in yield of crops (Vaux and Pruitt, 1983). Nonrecoverable deep percolation is another non-beneficial component of the total water depleted (CF) from thecrop land during irrigation (Allen et al., 1998). This also increases at higher dosage of irrigation, which occurin case of canal irrigation. Moreover, with controlled water delivery, the efficiency of utilization of fertilizerswould be more in the first case. Hence, with improved reliability and water delivery control, both denominato(CF) and numerator (yield) of water productivity parameter (kg/m3) could be higher. This can be better understoodby the negative correlation between surplus irrigation and crop yields in Sone command that surplus irrigationled to reduced yields (Meinzen-Dick, 1995). Since, there are no extra capital investments it would also lead tohigher productivity in economic terms.

The second possibility is that with greater quality and reliability of irrigation, the farmers are able toprovide optimum dosage of irrigation to the crop, controlling the non-beneficial evaporation, and non-recoverabledeep percolation, with the result that the CF remains low, and the fraction of beneficial evapo-transpirationwithin the CF or the depleted water remains high. Also, it is possible that with high reliability regime ofthe available supplies, even under scarcity of irrigation water, the farmers can adjust their sowing time suchthat they are able to provide critical watering. This can bring out high yield responses. Both result in higherWP in kg/ET.

But, does the differential WP in economic terms (Rs/m3) come from well owners growing more waterefficient and sensitive crop with assured water supplies? Evidence in support of this argument is a recent studycomparing water productivity of shareholders of tube well companies and water buyers in north Gujarat. Thestudy showed that the shareholders of tube well companies got much higher returns from every unit of pumpedwater, i.e., overall net water productivity in economic terms (Rs.4.18/m3), as compared to water-buyers (Rs.1.3m3). The reason was that water allocation for shareholders was quite assured in volumetric terms, and irrigationwater delivery was highly reliable, owing to which they could do their water budgeting properly, select watersensitive and high-valued crops, and make investments for inputs judiciously, whereas water buyers were at themercy of the well owners (Kumar, 2005).

Now, with expanding well irrigation in many arid and semi arid countries like India, including canacommand area, new opportunities for improvement in reliability of water supplies is available. If well irrigationgives positive differential WP over surface irrigation, we can build in such features that contribute to highewater productivity in well irrigation, in gravity irrigation systems. They include creating intermediate storagesystem for storing canal water; and lifting and delivery devices for the stored water. That said, in real economic

terms, what does the productivity gain means given the fact that the economic costs of irrigation is muchhigher than the private costs for both canal irrigation and well irrigation? Understanding these linkages will helpdesign better policies for water allocation (whether to supply water by gravity or promote conjunctive use) andpricing in surface irrigation. If reliability results in higher WP (Rs/m 3) in well irrigation, which cannot beexplained by price variations, then that makes tariff increase in canal water contingent upon improving thequality of irrigation.

3.1.2 Impact of changing water allocation on water productivity and water saving

Water management decisions are often taken on the basis of average water productivity estimates. Fothe same type of system, water productivity for the same crop can change at field scale (Singh et al., 2006:pp272


7/17

191

according to water application and fertilizer use regimes. Hence, it is important to know the marginal productivitywith respect to water and nutrient use. It helps to analyze the role of changing water allocation strategies at thefield level on enhancing WP. But, there are no data available internationally.

For a given crop, the irrigation dosage and the crop water requirement (beneficial use plus beneficianon-consumptive use) corresponding to the maximum yield may not correspond to the maximum wateproductivity (Rs/m3) (Molden et al., 2003). The WP (k/m3) would start leveling off and decline much before theyield starts leveling off (see Figure 1.2 in Molden et al., 2003). Ideally, WP in terms of net return from crop pecubic metre of water (Rs/m3) should start leveling off or decline even before physical productivity of water (kg

m3) starts showing that trend. When water is scarce, there is a need to optimize water allocation to maximizewater productivity (Rs/m3) through changing the dosage of irrigation. But, this may be at the cost of reducedyield and net return per unit of land, depending on which segment of the yield and WP response curves thecurrent level of irrigation corresponds to.

Recent analysis with data on applied water, yield and irrigation WP for select crops in the Narmadariver basin in India showed interesting trends. In many cases, trends in the productivity of irrigation water inresponse to irrigation did not coincide with the trends in crop yields in response to irrigation (Figure 1 andFigure 2); whereas in certain other cases the trends in irrigation WP in response to irrigation and the trends inyield in response to irrigation did actually coincide at least for some range in irrigation (Figure 3 and Figure 4)Knowing at what segment of the WP response curve irrigation dosage to a given crop lies helps understand how

changing water allocation would change the crop yield and WP.

y = 0.1759x + 1922.6

R2

= 0.14

0

1000

2000

3000

4000

0.0 1000.0 2000.0 3000.0 4000.0 5000.0 6000.0 7000.0 8000.0

Irr igation Dosage (m3/ha)

Yield(kg/ha)

y = -2.2153Ln(x) + 19.961

R2

= 0.28

0.00

2.00

4.00

6.00

8.00

0.00 1000.00 2000.00 3000.00 4000.00 5000.00 6000.00 7000.00 8000.00

Irrigation Dosage (m3/ha)

WaterProductivity

(Rs/m3)

Figure 1: Yield vs Irrigation Dosage in Wheat (Hoshangabad 2002)

Figure 2: Water Productivity vs Irrigation Dosage in Wheat(Hoshangabad 2002)


8/17

192

The regression values for the response of yield to irrigation dosage being very small (Figure 1 andFigure 3), one could argue that many factors other than irrigation explain yield variations. But, the data that arepresented here are for different farmers, who represent different soil conditions, different planting dates anddifferent seed varieties, all of which having a potential to influence the crop yield. If one takes into account thisone could say that the actual yield response to irrigation would be much stronger if planting date, soils and seedvarieties and same. Also, the slope of yield curve is very mild in the case of Figure 1. This is quite contrary towhat can normally be found given the wide range in irrigation water dosage among the sample farmers. Thiscan be explained by the variation in PET, and the moisture availability across farmers in the sample, which

changes the irrigation water requirements.

In the first case, where the level of irrigation corresponds to the ascending part of the yield curve, buthe descending part of WP curve (Figures 1 and 3), then limiting irrigation dosage might give higher net returnper unit of water. But, farmers may not be interested in that unless it gives higher return from the land. Henceif the return from the land does not improve, the strategy can work only under three situations: 1] the amountof water farmers can access is really limited either by the natural environmentlike limited groundwaterreserves; 2] there is a high marginal cost of using water due to high prices for water or electricity used forpumping water that it is much closer to the WP values at the highest levels of irrigation; and, 3] water supply irationed. In all these situations, the farmers should have extra land for using the water saved. Under conditionof supply rationing, farmers would anyway be using water for growing economically efficient crops. But, the

Figure 3: Yield vs Irrigation Water Dosage in Cotton (West Nimar 2003)

Figure 4: Water Productivity vs Irrigation Dosage in Cotton (West Nimar 2003)

y = -0.0001x2

+ 0.7534x + 322.54

R2

= 0.2998

0.00

500.001000.00

1500.00

2000.00

2500.00

0.00 1000.00 2000.00 3000.00 4000.00 5000.00 6000.00

Irrigation Water Dosage (m3 /ha)

Yield(Kg/ha)

y = -5E-07x2+ 0.0025x + 0.6815

R2

= 0.15

-2.00

0.00

2.00

4.00

6.00

8.00

10.00

0.00

1000.00 2000.00 3000.00 4000.00 5000.00 6000.00

Irrigation Dosage (m3/ha)

WP(Rs/m3)


9/17


10/17


11/17

195

(Kumar and Singh, 2001). Hence, water rationing is important to affect demand regulations in most situations(Perry, 2001a). The challenge lies in understanding the science of WP, particularly WP response to irrigationand actual consumptive use of water, and managing irrigation water deliveries accordingly. In the case of welirrigation, it is important for the farmers to understand this linkage, whereas the official agencies have to ensurethat power supply is available for critical waterings.

As regards the third point, often the farmers do not make correct judgments about the level of irrigationdosage that corresponds to zero marginal returns. This has been found in the case of well owners, who are noconfronted with positive marginal cost of pumping, resulting in lowering yield with incremental irrigation

(Kumar, 2005). Price reforms only make farmers more conscious about the negative economic consequencesof giving over-dose of irrigation water.

3.2 Opportunities for Improving Farm-level and Regional Level Water Productivity

We have seen that there are clear trade offs between options to enhance physical productivity of wateand WP in economic terms at the field level itself. We would see that there is trade off between maximizing WPat the field level and that at the farm level, though farm level water productivity is dependent on the processethat govern WP at the individual fields. We would also see that the options available to maximize WP in a regionwhich often is the concern of water policy makers, are much less than those for an individual farms. The watepolicy maker looks for approaches that would not only enhance the economic returns, but also increase thesocial welfare. Many of the decisions relating to public investment in irrigation systems in countries like Indiaare driven by larger societal concerns such as producing more food, employment generation and povertyalleviation. Often policy makers are more driven by social and political considerations than purely economicconsiderations (Perry, 2001a). We would elaborate on these issues in the subsequent paragraphs.

From the analysis presented in the previous section, it is evident that the scope for improving field leveWP is extremely limited given the social, economic, institutional and policy environment in India. Limitations aremore when we want to use it as a driver for changing water demand. Therefore, WP enhancement should

1. Kharif Paddy 7.75 4.78 -

2. Fodder Bajra 2.93 4.78 -

3. Kharif Cotton 40.40 - -

4. Kharif Castor - - 8.09

5. Brinjal - - -

6. Wheat 8.05 9.11 4.46

7. Fodder Jowar 6.32 -8. Mustard - - 4.73

9. Winter Gram 24.48 - -

10. Jowar - - 4.01

11. Cumin - - 19.84

12. Summer Bajra - - 2.85

Source: based on Kumar et al. (forthcoming) for western Punjab and eastern UP; and Kumar (2005) for northGujarat. In the case of north Gujarat crops, the mean values of water productivity figures for differentcategories of farmers were taken.

Table 3: Applied Water Productivity in Selected Crops in North Gujarat, Western Punjab and easternUttar Pradesh

Sr.No.

Name of the CropNet Water Productivity of Crop (Rs/m3) of Applied Water in

Western Punjab Eastern UP North Gujarat


12/17

196

focus on crops that are inherently more water efficient in economic terms, but also have high return per unit oland. As Molden (per. com) notes, increasing WP is not often relevant to farmers. If it is important to thesociety, then society should figure out ways to align everyones incentives.

It is established that many fruit crops have higher WP (Rs/m 3) than the conventional cereals such awheat and paddy in arid areas. For instance, pomegranate grown in north Gujarat gives a net return of nearly40,000 rupees per acre (i.e., USD 1000/acre) of land against Rs. 8,000 per acre (i.e., USD 200/acre) in case owheat. The WP is approximately Rs.100/m3 for pomegranate (Kumar, 2007) against Rs. 4.46/m3 for wheat inthe same region. Also, there are crops such as potato, cumin, cotton and castor which are more water efficien

than rice and wheat, which can be grown in Punjab (see Table 3). With greater reliability, and control over watedelivery, farmers using well irrigation would allocate more water for growing water-efficient crops. Perhapsfarmers have already started shifting to high valued cash crops.

But, there are limits to the number of farmers who can take up such crops due to the volatile nature othe market for most of these crops, its perishable nature, and the high risk involved in producing the crop. Forinstance, cumin grown in north Gujarat is a very low water consuming crop, with a high return per ha. Butcrop failure due to disease is very common in cumin. In case of vegetables, that are fast perishable, markets areoften very volatile, and price varies across and within seasons. The problem of price fluctuation is also applicableto cotton grown in western Punjab, which has high WP. Also, the investments for crops are also very highdemanding risk-taking ability.

But, farmers organize their entire farm, rather than field to maximize the net economic return(Ruthenberg, 1980). The extent to which farmers can allocate water to economically efficient crops wouldperhaps be limited by the need to manage fodder for animals. It may also get limited by the poor market supporfor orchard crops. Many farmers in Punjab and other semi arid parts of India, manage crops and dairy farmingtogether. But, even globally, research analyzing WP in composite farming systems that really take into accounwater depleted in biomass production is almost absent. Literature on water use efficiency and WP in dairyfarming is also extremely limited. In regions for which they are available, the conditions are extremely differentfrom that in countries like India. Studies from northern Victoria and Southern New South Wales analyzed wateuse efficiency in dairy farms that are irrigated (Armstrong et al., 2000) and dairy farming is not integrated withcrop production in this region. Green fodder produced in irrigated grass lands is used to feed the cattle by dairyfarmers in Australia and United States, unlike Sub-Saharan Africa and developing countries in south Asia.

Recent analyses from western Punjab seem to suggest that the overall net WP in rupee terms getsenhanced when the byproducts of cereal crops are used for dairy production (see Table 4).

Sr. No Name of Crop/Farming Water Productivity (Rs/m3)

1. Paddy 7.75

2. Wheat 8.05

3. Milk Production 13.06

Source: Kumar et al., forthcoming (derived from Table 11)

Table 4: Water Productivity in Crops and Dairy Production

Reduced area under cereal crops such as paddy and wheat would mean reduction in availability ofodder. Farmers may have to grow special crops that give green fodder, and in that case, they might in turn beincreasing the water use intensity7. Otherwise, farmers may have to procure dry fodder from outside, whichwould involve more labour. Hence, there could be a trade off between maximizing crop WP and farm levelWP. But, there is not much of literature about economic productivity in dairy farming, especially with cerealsand dairying, to understand this trade off.

7 In a similar semi-arid situation in north Gujarat, it was found that dairy production, which used irrigated alfalfa, was highly waterinefficient, both physically and economically (Singh, 2004).


13/17

197

At the regional level, enhancing WP through either shift to water efficient crops (like orchards andvegetables) or with crop-dairy based farming system might face several constraints from socio-economic poinof view. Food security is an important consideration when one thinks about options to enhance WP. Labourabsorption capacity of irrigated agriculture and market price of fruits are other considerations. Paddy is labouintensive and in fact a large chunk of the migrant labourers from Bihar work in the paddy fields of PunjabReplacing paddy by cash crops would mean reduction in farm employment opportunities. On the other handthe lack of availability of labour and fodder would be constraints for intensive dairy farming to maximize WP athe regional level, though some farmers might be ale to adopt the system. Large-scale production of fruits mighlead to price crash in the market, and farming loosing revenue unless sufficient processing mechanisms areestablished. Hence, the number of farmers who can adopt such crops is extremely limited.

In a developing country context, poverty reduction potential or the food security impact of irrigationare more important than return per unit of water. Food security and poverty reduction are in-built goals in largescale subsidies in irrigation (Gulati, 2002), which enable poor farmers to intensify cropping. Therefore, WP inirrigation needs to be looked at from that perspective also, and not merely crop per drop. One can argue thawith more reliable irrigation, farmers could as well produce more food or generate more employment, and withthat achieve higher physical and economic productivity along with meeting social objectives. But, the heavysubsidies in irrigation reduce the ability of the agencies to improve its quality through regular investments.

Perhaps this welfare oriented policy of keeping irrigation charges low now needs a re-look. With

extensive well irrigation in India and with the poor paying heavy charges for pump renting or well water toirrigate their crops, the policies to subsidize canal irrigation may not bring about the desired equity and welfareoutcomes. In fact, a large chunk of the subsidies in canal irrigation goes to large farmers, due to the crop-areabased pricing followed (Kumar and Singh, 2001). These farmers also have access to well irrigation in thecommand area.

Another fact that supports the above argument is that often the unreliable canal water supplies forcefarmers to adopt only paddy, and not domestic food security concerns. The stable and high procurement priceoffered by the Food Corporation of India for cereals such as rice and wheat allow farmers to stick to thiscropping system. But there are major macro economic imperatives of trying to meet these social objectives(Gulati, 2002). The intensive paddy cultivation in Punjab is associated with intensive use of electricity fo

pumping groundwater even in canal commands during summer. Irrigating one ha of winter wheat requires 74Kwhr to 295 Kwhr of electricity, which costs Rs.300 to Rs.1175 to the economy (source: field data). Theregion is already facing power crisis, with resultant impact on the quality of power supply to farm sectorEnhancing productivity of pumped groundwater also means enhancing energy productivity and reducing therevenue losses to the government in terms of power subsidies.

If farmers are able to secure higher net return from every unit of water applied or depleted in welirrigation, this could be a major starting point for irrigation bureaucracies to start charging higher for irrigationalong with improving the qualityadequacy, reliability and control. Following norms of rationing in waterallocation would be crucial in achieving higher WP. Perhaps, what would be required would be higher prices forfood crops or special incentives for farmers who grow it so as to reflect its social benefit, while reducing the

irrigation subsidies heavily. So, the net result would be a compromise between socio-economic productivityand productivity enhancement in monetary terms, with positive impact on the water resource system.

4. SUMMARY AND CONCLUSIONS

Research to explore potential improvements in physical productivity of water (kg/ET) in crops withoudue consideration to income returns per unit of water will not be relevant for Indian farmers under the currentelectricity and water pricing policies in agriculture, and institutional regimes governing water use. The reason iit does not link WP improvement to raising aggregate farm income. In countries like India, major determinantfor analyzing improvements in basin level WP due to WH & supplementary irrigation should use: i] incrementaeconomic returns from enhanced crop yield; and ii] opportunity costs of water harvesting at basin scale


14/17

198

Analysis of basin level impacts of efficient irrigation technologies on basin WP and water saving should involveconsider CF as a determinant of WP rather than evapo-transpiration.

Research on potential impact of improved reliability of irrigation water and changing water allocationon WP is relevant for developing countries like India as it gives due consideration to maximizing farmersincome, while reducing the total water depleted. Nevertheless, their overall potential in improving WP inagriculture and more so in reducing water demand is open to question, unless policies and institutions arealigned to make societys interests and farmers interests match. For the composite farming systems that arecharacteristic of countries like India, WP research should focus on optimizing water allocation over the entire

farm to maximize the returns, through changes in crop mix and crop-livestock compositions. But, dueconsideration should be given to risk taking ability of the farmer, investment capabilities etc.

In countries like India, research on measures to enhance regional level WP should integrate socioeconomic considerations such as food production, employment generation along with wealth generated per uniof water used up in irrigation. But, often farmers choice of food crops like rice is not by design, but by defaultMeeting food production needs or other social objectives cannot be an excuse for poor productivity. Giventhese constraints, regional WP scenarios can examine the scope for improving WP through increment inproductivity of crops such as wheat and paddy with reliability and control regimes in irrigation, along with othemeasures.

To conclude, the options to enhance WP in crops in countries like India seem to be quite limited, and

different from those being tried in the West, given the larger objective of addressing food security, povertyalleviation, and employment generation concerns in rural areas. Research should aim at strategies to enhanceWP that are based on improving reliability, adequacy, and water allocation for reducing non-beneficial consumptiveuse, and non-beneficial non-reusable portions of water supplies. The inherent advantages of well irrigationsystems need to be built in while designing surface irrigation systems and designing water allocation normsBut, in most cases, they could regulate water demand only if water allocation is rationed volumetrically.

ACKNOWLEDGEMENT

The authors are highly thankful to Dr. Hugh Turral, Dr. David J. Molden, Dr. Charlotte De Fraiture andDr. Mobin-ud-din Ahmad, Principal Researchers, International Water Management Institute, and Dr. Petra

Hellegers, Associate Professor of Economics at Wageningen University and Research Centre, for their views onvarious determinants and drivers of change in water productivity in agriculture. They were of immense help indeveloping several of the arguments presented in this paper.

REFERENCES

ABS (2002), Census 2001, Australian Bureau of Statistics, Canberra.

Ahmad, M. D., I, Masih and H. Turral (2004), Diagnostic analysis of spatial and temporal variations in cropwater productivity: A field scale analysis of the rice-wheat cropping system of Punjab, Pakistan, Journaof Applied Irrigation Science, 39 (10).

Ambast, S. K., N. K. Tyagi, S. K. Raul (2006), Management of declining groundwater in the Trans IndoGangetic Plain (India): Some options, Agricultural Water Management, 82 (2006): 279-296.

Armstrong, D. P., J. E. Knee, P.T. Doyle, K. E. Pritchard and O. A. Gyles (2000), Water-use efficiency onirrigated dairy farms in northern Victoria and southern New South Wales, Australian Journal ofExperimental Agriculture, 40: 643-653.

Allen, R. G., L. S. Willardson, and H. Frederiksen (1998), Water Use Definitions and Their Use for Assessingthe Impacts of Water Conservation in: (J. M. de Jager, L.P. Vermes, R. Rageb (Eds), Proceedings ICIDWorkshop on Sustainable Irrigation in Areas of Water Scarcity and Drought Oxford, England, September

11-12, pp 72-82.


15/17

199

Chakravorty, Ujjayant and Chieko Umetsu (2003), Basinwide water management: a spatial model, Journalof Environmental Economics and Management, 45 (2003): 1-23.

Bosworth, B., G. A. Cornish, Chris J. Perry and F. van Steenbergen (2002), Water charging in irrigated agriculturelessons from the literature. HR Wallingford Ltd. Wallingford, UK. OD145.

Brewer, J., Shashi Kolavalli, A.H. Kalro, G. Naik, S. Ramnarayan, K.V. Raju and R. Sakthivadivel (1999)Irrigation Management Transfer in India: Policies, Processes and Performance. Oxford & IBHNew Delhi and Calcutta.

Connor, D. J., and T. R. Jones (1985), Response of sunflower to strategies of irrigation. II. Morphological andphysiological responses to water shortage. Field Crops Research, 12: 91103.

De Fraiture, Charlotte and Chris Perry (2002), Why is irrigation water demand inelastic at low price ranges?Paper presented at the conference on irrigation water policies: micro and macro considerations, 15-17June 2002, Agadir, Morocco.

Falkenmark, Malin and Johan Rockstrom (2004), Balancing water for human and nature: a new approach ineco-hydrology, USA and UK: Earthscan.

Fox, P., J. Rockstrom and J. Barron (2002), Risk Analysis and economic viability of water harvesting fo

supplementary irrigation in semi arid Burkina Faso and Kenya, Agricultural Systems, 83 (3): 231-250

GOI (1999), Integrated Water Resource Development A Plan for Action, Report of the National Commission onIntegrated Water Resources Development, Volume I, Ministry of Water Resources, Government of IndiaNew Delhi.

Grismer, E. M. (2001), Regional cotton lint yield and water value in Arizona and California. Agricultural WateManagement, 1710, 1-16.

Gulati, Ashok (2002), Challenges to Punjab Agriculture in a Globalizing World, paper based on the presentationgiven at the policy dialogue on Challenges to Punjab Agriculture in a Globalizing World jointly organizedby IFPRI and ICRIER, New Delhi.

Haisman, B. (2003), Impacts of Water Rights Reform in Australia, International Working Conference on WaterRights, International Food Policy Research Institute.

Hernndez-Mora, N., Llamas, M.R. and Martnez, L. (1999), Misconceptions in aquifer overexploitationimplications for water policy in Southern Europe, 3rd Workshop SAGA. Milan, December (1999) Preprin22 p.

Howell, Terry (2001), Enhancing Water Use Efficiency in Irrigated Agriculture, Agronomy Journal (93): 281289.

Institute of Rural Management Anand/ UNICEF (2001), White Paper on Water in Gujarat, Report prepared forthe Government of Gujarat.

Loomis, R. S. and D. J. Connor (1996), Crop ecology: Productivity and Management in Agricultural SystemsCambridge, New York and Australia, Cambridge University Press.

Johansson, R. C. (2000), Pricing Irrigation Water: A Literature Survey. World Bank, Washington, D. C. 80pp

Kijne, Jacob, R. Barker and D. Molden (2003), Improving Water Productivity in Agriculture: Editors Overviewin Jacob Kijne and others (Eds.) Water Productivity in Agriculture: Limits and Opportunities foImprovement, Comprehensive Assessment of Water Management in Agriculture. UK: CABI Publishingin Association with International Water Management Institute.


16/17


17/17

Perry, Chris J. (2001a.), Water at any Price? Issues and Options in Charging for Irrigation Water, Irrigation andDrainage, 50 (1): 1-7.

Perry, Chris J. (2001b.), World Commission on Dams: Implications for Food and Irrigation, Irrigation andDrainage, 50: 101-107.

Ray, Isha (2002), Farm level incentives for irrigation efficiency: some lessons from an Indian canal. (Draft)

Ray, Sunil and Bijarnia (2006), Upstream vs Downstream: Groundwater Management and Rainwater Harvesting

Economic and Political Weekly, July 10, 2006.Rijsberman, Frank R. (2004), Copenhagen Consensus Challenge Paper, The Water Challenge, Copenhagen

Consensus 2004.

Rockstrm, J., Jennie Barron and Patrick Fox (2002), Rainwater management for improving productivityamong small holder farmers in drought prone environments, Physics and Chemistry of the Earth, 27(2002): 949-959.

Ruthenberg, H. (1980), Farming systems of the tropics, Clarendon Press, Oxford.

Scott, Christopher and Silva-Ochoa (2001), Collective action for water harvesting irrigation in Lerma-Chapala

basin, Mexico, Water Policy, 3: 55-572.Seckler, David (1996), New era of water resources management: from dry to wet water savings, Research

Report 1, International Irrigation Management Institute, Colombo, Sri Lanka.

Seckler, David, D. Molden and R. Sakthivadivel (2003), The Concept of Efficiency in Water ResourcesManagement, in Jacob Kijne and others (Eds.) Water Productivity in Agriculture: Limits and Opportunitiefor Improvement, Comprehensive Assessment of Water Management in Agriculture. UK: CABI Publishingin Association with International Water Management Institute.

Singh, O. P. (2004), Water-intensity of North Gujarats Dairy Industry: Why Dairy Industry Should Take aSerious Look at Irrigation?, paper presented at the second International Conference of the Asia Pacific

Association of Hydrology and Water Resources (APHW 2004), Singapore, June 5-9, 2004.

Singh, Ranvir (2005), Water productivity analysis from field to regional scale: integration of crop and soimodeling, remote sensing and geographic information, doctoral thesis, Wageningen UniversityWageningen, The Netherlands.

Thobani, Mateen (1997), Formal water markets: why, when, and how to introduce tradable water rights. WorldBank Research Observer, 12 (2), August 1997, 161-79.

Tsur, Y and Arial Dinar (1995), Efficiency and Equity Considerations in Pricing and Allocating Irrigation WaterPolicy Research Working Paper 1460. The World Bank. 40pp.

Van Dam, Jos, R. Singh, J.J. E. Bessembinder, P. A. Leffelaar, W. G. M. Bastiaanssen, R. K. Jhorar, J.C. Kroesand P. Droogers (2006), Assessing options to Increase Water productivity in Irrigated River BasinUsing Remote Sensing and Modelling Tools, WaterResources Development, 22 (1): 115-133

Vaux, H. J. Jr., and W. O. Pruitt (1983), Crop-water production functions, in Advances in Irrigation, volume 2

ed. D. Hillel. Orlando, Florida, USA: Academic Press.

Zwart, Sander J. and Wim G. M. Bastiaanssen (2004), Review of measured crop water productivity values foirrigated wheat, rice, cotton and maize, Agricultural Water Management 69 (2004): 115-133.

Documents

Improving Water Productivity-new Avenues (IWMI-TATA)